Manufacturing facilities often utilize real-time monitoring to ensure that various pieces of equipment are operating within allowable tolerances. In semiconductor manufacturing facilities in particular, improving the control of a fleet of manufacturing tools to act interchangeably has become increasingly important in the face of 3D metrology and tight design rules. Each tool of the fleet may perform measurements in real-time to ensure that measurements of critical dimension of features are within acceptable ranges. Traditional methods have focused on improving upon the measurement uncertainty on dedicated monitoring artifacts. In doing so, however, such methods are more sensitive to the dedicated monitoring artifacts rather than to the performance of the fleet on production artifacts, and thus lack the ability to identify excursions early on during steady state fleet operation.